I am adding a PS to my post. Vioxx was yanked by Merck because a pain killer was shown in a RCT to have a relative risk of MI events of 1.92 (rofecoxib 1.50 events vs placebo 0.78 events per 100 patient years). The results from the first 18 months of the APPROVe study did not show an increased relative risk of adverse cardiovascular events. All the problems cropped up after. And still both the FDA and Canadian Health advisory panels recommended returning Vioxx to the market for many patients under some conditions. In each case panelists believed that uncontrolled pain was a greater risk than the absolute risk of a heart attack from long term use associated with any NSAID.
In Avandia's case, we have no clear cut evidence of MI in any long term data except in the insulin population that doctors tend not to give Avandia to in the first place. So for the FDA to say Monday's meeting is not about Nissen is misleading. It is all about Nissen. It would not be meeting July 30 if not for his end run with his "study." More to the point, the FDA the failed to ask if uncontrolled glucose levels was a greater risk than the absolute risk of a heart attack from long term use associated with any oral diabetes drug.
That's the Nissen effect.